
Gmail
io.github.domdomeggPublished by GitHub account domdomeggv2.4.0Updated Oct 24, 2026
Allow AI systems to read, send, archive, and manage emails via Gmail.
Overview
Local MCP server that lets an assistant read, search, send, archive and manage Gmail messages, threads, drafts, labels, filters and vacation settings.
- What you can do
- Read profile info and search or list messages and threads, then fetch full threads or single messages.
- Send, forward, label, archive, trash, restore or permanently delete messages and threads, including bulk operations.
- Create, update, send and delete drafts.
- Manage labels and filters, and download attachments.
- Read and set vacation auto-reply settings.
- When to use it
- Useful when an assistant should work inside a Gmail inbox: finding confirmations, receipts or attachments, drafting and sending replies, tidying mail with labels and archives, or turning on an auto-reply while away.
- Requirements
- Runs locally via npx (Node.js). Needs Google Cloud OAuth credentials with the Gmail API enabled: GOOGLE_CLIENT_ID, GOOGLE_CLIENT_SECRET and GOOGLE_ACCESS_TOKEN, plus MCP_TRANSPORT. The README describes running it as an HTTP server on localhost:3000 with a Google OAuth redirect URI registered.
Installation
In SourceWeft
- Open Gmail in the dashboard and add it to a workspace.
- Enable the server for the chats that should use its tools.
Desktop only via STDIO. STDIO servers start a local process, so they need the SourceWeft desktop host.
Other MCP clients
Follow the launch instructions in the repository.
README
gmail-mcp
MCP server for Gmail - read, send, archive, and manage emails.
Use Cases
Flight check-in: "Check me in for my flight tomorrow" → finds your confirmation email, extracts the booking reference, and completes check-in with browser use.
Expense submission: "Submit my Amazon purchases from last week as expenses" → finds receipts, extracts amounts/categories, and files them in your expense system.
Package tracking: "Where's my package from Acme?" → finds the shipping notification, extracts the tracking number, and fetches current delivery status.
Find that attachment: "What did we settle on for the budget? Michelle sent me something yesterday" → locates the email and opens the attached spreadsheet.
Vacation mode: "I'm on holiday Dec 20-Jan 2" → sets up Gmail auto-reply, blocks your calendar, and updates your Slack status in one go.
Meeting follow-up: After a call, your assistant reads the Gemini transcript, summarizes key decisions, and drafts a follow-up email with action items.
(These are just examples - any workflow that needs email search, reading, or sending can use this.)
Setup
1. Create Google OAuth credentials
- Go to Google Cloud Console
- Create a new project (or use existing)
- Enable the Gmail API
- Go to APIs & Services → OAuth consent screen, set up consent screen
- Go to APIs & Services → Credentials → Create Credentials → OAuth client ID
- Choose Web application
- Add
http://localhost:3000/callbackto Authorized redirect URIs - Note your Client ID and Client Secret
2. Run the server
The server runs on http://localhost:3000 by default. Change with PORT=3001.
3. Add to your MCP client
With the server running, follow the instructions on install-mcp, which generates the right config for your MCP client (Claude Code, Claude Desktop, Cursor, Cline, VS Code, and more).
Architecture
This server acts as an OAuth proxy to Google:
- Server advertises itself as an OAuth authorization server via
/.well-known/oauth-authorization-server /registerreturns the Google OAuth client credentials/authorizeredirects to Google, encoding the client's callback URL in state/callbackreceives the code from Google and forwards to the client's callback/tokenproxies token requests to Google, injecting client credentials/mcphandles MCP requests, using the bearer token to call Gmail API
The server holds no tokens or state - it just proxies OAuth to Google.
Tools
Gmail API Scopes
gmail.readonly- Read messages and labelsgmail.send- Send messagesgmail.modify- Modify messages (archive, labels, trash)
Contributing
Pull requests are welcomed on GitHub! To get started:
- Install Git and Node.js
- Clone the repository
- Install dependencies with
npm install - Run
npm run testto run tests - Build with
npm run build
Releases
Versions follow the semantic versioning spec.
To release:
- Use
npm version <major | minor | patch>to bump the version - Run
git push --follow-tagsto push with tags - Wait for GitHub Actions to publish to the NPM registry.
Source: README.md at commit 32103e7
Tools
0Version history
1- v2.4.0LatestSep 16, 2026
Score
40% of the applicable checks met
How complete this listing is, computed from the registry entry, the endpoint and public sources. Not a rating of the server and not a statement that it is safe or working today: reachability is a dated check, and no installs or reviews go into the score.
- Reached in the last 30 days25 pts
Keep the endpoint answering the MCP handshake; SourceWeft probes it every 30 days.
- Auth declared10 pts
Declare the credential headers in the registry entry, or keep the endpoint answering the probe so its auth can be classified.
- Tools listed15 pts
Answer tools/list without a credential, or ship a package that starts in a sandbox without one.
- DescriptionRequired15 pts
- Logo10 pts
- Documentation15 pts
- License10 pts
Declare a license in the repository or the registry entry.